Munich Personal RePEc Archive

Determinants of Financial Inclusion in

Africa: A Dynamic Panel Data Approach

Evans, Olaniyi

Pan Atlantic University, Lagos, Nigeria

2016

Online at https://mpra.ub.uni-muenchen.de/81326/
MPRA Paper No. 81326, posted 13 Sep 2017 08:43 UTC
 UNIVERSITY OF MAURITIUS RESEARCH JOURNAL – Volume 22 – 2016
University of Mauritius, Réduit, Mauritius

DETERMINANTS OF FINANCIAL INCLUSION IN AFRICA:

A DYNAMIC PANEL DATA APPROACH

Olaniyi Evans
Department of Economics, University of Lagos, Akoka Lagos, Nigeria
&

*Babatunde Adeoye, Ph.D

Department of Economics, University of Lagos, Akoka Lagos, Nigeria
Email: badeoye2010@yahoo.com

Abstract
This study documents the determinants of financial inclusion in Africa for the period 2005 to
2014, using the dynamic panel data approach. The study finds that per capita income, broad
money (% of GDP), literacy, internet access and Islamic banking presence and activity are
significant factors explaining the level of financial inclusion in Africa. Domestic credit provided
by financial sector (% of GDP), deposit interest rates, inflation and population have insignificant
impacts on financial inclusion. The findings of this study are of utmost value to African central
banks, policymakers and commercial bankers as they advance innovative approaches to enhance
the involvement of excluded poor people in formal finance in Africa.

Keywords: Financial inclusion, finance, dynamic panel data, Africa

JEL Classification: C23, E62, F30, D14, G21

 O. Evans and B. Adeoye

Introduction

What are the determinants of financial inclusion in Africa? Recently, there has been a rapid
thrust for financial inclusion, more so in emerging economies, such as in Africa. However,
financial inclusion continues to pose increasing concerns for a vast number of economies. The
concerns have led more than 50 countries to set formal targets of universal financial access by
2020 (i.e. Lesotho, Nigeria, Rwanda) and many more countries tasking their regulatory and
supervisory agencies with encouraging financial inclusion. These increasing concerns are
necessary considering the possible “cost in foregone economic growth when the volume of
financial services in a country does not reach a sufficiently large share of the population”
(Barajas, Chami & Yousefi, 2013 as cited in Naceur et al, 2015, p.4).

The share of unbanked adults can be as high as 90%, in many emerging and developing
economies. Among the emerging and developing economies, account ownership is lowest in
Africa (Mehrotra & Yetman, 2015). However, the level of financial inclusion varies widely by
country and income level. For example, the adult population having a bank account is highest in
Mauritius (80%) and South Africa (54%), followed by Angola, Mozambique, Kenya,
Zimbabwe, and Morocco (all around 40%). With its mobile banking leading the way, Kenya has
a successful financial inclusion policy. In addition, 14 African countries have less than 10% of
the adult population having a bank account with a formal financial service (i.e. Egypt, Guinea,
Niger, and Congo). For instance, in the Central African Republic, Kinshasa, Guinea and Congo,
less than 5% of the adult population have access to formal financial services. In Niger, only 2%
of the population has a bank account (Demirgüç-Kunt & Klapper, 2012).

Considering the character of finance, two types of factors constitute financial inclusion
constraints in Africa. One type is time-invariant (structural characteristics) and the second type is
policy factors (Naceur et al, 2015). The structural characteristics are population size and density,
income per capita, and the level of economic informality, which all jointly decide the financial
market size. The policy factors are the degree of macroeconomic stability and the institutional
frameworks which determine the costs and risks of financial services provision. Thus, financial
inclusion can be seen as a function of both structural and policy factors.

2
 O. Evans and B. Adeoye

Naceur et al (2015) used an innovative graphical example to illustrate how financial inclusion
can be seen as a function of both structural and policy factors. Let’s consider the financial
inclusion proxy introduced in this study: depositors with commercial banks.1 The higher the
level of the economy, its population and density, the more cost-effective is loan provision to the
broader population. Therefore, as in Figure 1, a positive relationship is expected between this
access (on the vertical axis) and the structural characteristics (on the horizontal axis). As shown
by the structural access line, Country A, a small low-income country with a highly dispersed
population (depicted as STRUCTA) would be expected to have a lesser level of inclusion than
Country B, a large high-income country with concentrated urban population (depicted as
STRUCTB).

Figure1. Financial Possibility Frontier

Source: Naceur et. al (2015, p. 18)

1 A similar analysis could be carried out on the other financial inclusion variable: broad money per capital.

3
 O. Evans and B. Adeoye

However, pro-finance policies and better enabling environment can push Country A to break
through its structural constraints to register DA level of inclusion (DA> SDA). On the contrary,
country B, though with a higher level of access than country A may underperform compared to
countries with comparable structural characteristics (DB < STRUCTB). Therefore, well-thought-
out pro-finance policies and better enabling environment can drive access upward for African
countries with diverse structural characteristics, to eventually reach a maximum sustainable
level, termed by Naceur et al (2015) as the Financial Possibility Frontier (FPF) for inclusion.

In spite of the different conceptualizations of the factors that influence financial inclusion, little
effort has been made to empirically determine these determinants of financial inclusion or factors
that foster inclusion, especially in Africa. Existing studies have been devoted to the measurement
and promotion of financial inclusion, to the detriment of the empirical evaluation of its
determinants and impacts. This study is therefore a novel effort: it contributes to our
understanding of the determinants of financial inclusion in Africa, on which there is little
empirical research to date. This study therefore fills the gap by evaluating the determinants of
financial inclusion in Africa, using dynamic panel data approach for 15 countries over the 2005-
2014 period.

Regarding the motivation for this study, we selected Africa for quite a few reasons. Firstly, there
is a shortage of research on the determinants of financial inclusion in Africa. Secondly, among
the emerging and developing economies, account ownership is lowest in Africa (Mehrotra &
Yetman, 2015. The determinants of financial inclusion may therefore be more significant and
easily evident in Africa than in other parts of the world. Thirdly, many researchers consider that
financial inclusion has had significant benefits for economic development worldwide. Fourthly,
Africa has 34 of the current 48 Less Developed Countries worldwide, meaning the continent is a
splendid context for the evaluation of the determinants of financial inclusion.

The rest of the article is organized as follows. Section 2 briefly provides the literature review.
Section 3 is an exposition of data and methodology. Section 4 describes the results of the
dynamic panel analysis. Section 5 concludes with overall summary and policy suggestions.

4
 O. Evans and B. Adeoye

Literature Review

The development of the financial sector, together with financial liberalization and international
financial integration, came to the epicenter of academics as well as policy-makers discussions
during the last four decades (Mirdala, 2011), due to its potential effects on the general economic
performance of an economy. As a matter of fact, in the early neoclassical growth literature, the
financial sector only served as a conduit of household savings to investors. However, studies
such as Goldsmith (1969) and Mickinnon (1973) had a contrary view, proposing a more robust
character for financial services. Ever since, a considerable volume of empirical and theoretical
literature has looked at the role of an efficient financial system as the foundation for sustaining
an open, vibrant economic system (i.e. Demetriades & Andrianova, 2004; Godhart, 2004; Abu-
Bader & Abu-Qarn, 2008; Mirdala, 2011; Adeoye and Saibu, 2014; Adeoye & Sangosanya,
2015 and Adeoye, 2015 ).

The success of the financial system in contemporary times is attributed to financial sector
reforms such as market-based procedures, the promotion of competition, and the relaxation of
restrictions. The goal of these reforms is to initiate a more efficient and stable system, which will
provide a foundation for effective stabilization and mobilization of capital, which leads to
reduced poverty (Johnston & Sundararajan, 1999; Adeoye and Saibu, 2014; Adeoye &
Sangosanya, 2015 and Adeoye, 2015). Emerging markets attach great importance to the
development and deepening of the financial sector in the pursuit of their poverty-reduction
objectives and, most recently, financial inclusion. Through mobilization of savings, facilitation
of payments, and promotion of proficient allocation of resources, the financial sector plays a
crucial role in the broadening of access to finance, stemming the tide of poverty (Zhuang,
Gunatilake, Niimi, Khan, Jiang, Hasan, & Huang, 2009) and thus, leading to increased financial
inclusion.

Of all the vast definitions of financial inclusion, the World Bank 2014 Global Financial
Development Report’s definition of financial inclusion as “the share of the population who use
financial services” is the most “useful definition because it can be measured and incorporated
easily into theoretical and empirical work” (Naceur, et al, 2015, p. 4). Thus, financial inclusion

5
 O. Evans and B. Adeoye

describes a situation where the bulk of financial services in a country reach a sufficiently large
share of the population.

As financial inclusion is becoming increasingly important for a vast number of countries

worldwide, a growing literature has been evaluating its measurements, determinants, and effects.
Bhattacharaya and Wolde (2010) found that inadequate access to credit has been undermining
growth in MENA countries compared to the rest of the world. Studies such as Naceur, et al
(2015) has identified a number of determinants of financial inclusion. First are structural factors.
Structural factors determine the costs of financial services provision to the population. Second
are policy-related factors. Policy-related factors can create enabling environments for financial
inclusion. For example, Love and Martínez Pería (2012) found that countries with more
competitive banking systems can enjoy greater financial inclusion as a higher number of firms
will have access to a loans and overdrafts. Third are some non-policy characteristics of the
country. For example, Aga and Martinez Pería (2014) show that inflows of international
remittances to sub-Saharan Africa play a part as well since remittance recipients are more than
likely to have a bank account.

A stream of research has shown the significant beneficial effects of financial inclusion for
individuals, thus supporting the economic and political rationale for financial inclusion-
promoting policies. For example, studies such as Banerjee and Newman (1993), Galor and Zeira
(1993), Aghion and Bolton (1997) and Beck, Demirguc-Kunt and Levine (2007) have
established an array of models to demonstrate how inadequate access to finance can lead to
inequality and poverty traps. The literature, at the same time, has established that access to
savings instruments increases consumption (Dupas & Robinson, 2009), savings (Aportela, 1999),
productive investment (Dupas & Robinson, 2009) and female empowerment (Ashraf, Karlan and
Yin, 2010). The literature further documents the significance of financial inclusion (see
Sherraden & Grinstein‐Weiss, 2015; Naceur, Barajas & Massara, 2015; Arya, 2015; Sarka, 2015;
Adeoye & Sangosanya, 2015 and Adeoye, 2015), claimed to be associated to necessary
economic outcomes beyond those linked to the more popular concept of financial depth.

The literature has established that while there is involuntary financial self-exclusion, there is, as
well, voluntary financial exclusion (de Koker & Jentzsch 2011). The voluntarily self-excluded

6
 O. Evans and B. Adeoye

decline to use financial services because they have no need of them or for cultural or religious
reasons (Beck, Demirgüç-Kunt and Honohan, 2009). The voluntarily excluded lacks trust in the
financial system (Dittus and Klein, 2011) or faces barriers such as affordability, inappropriate
product design and inability to meet eligibility criteria (European Commission 2008).

Apart from the voluntarily self-excluded and the involuntary self-excluded are the self-
withdrawn. The self-withdrawn are the bank customers who scaled access barriers but, then,
withdraw from the financial system. For example, the FinScope (2004) found that, in South
Africa, more than 3.5 million customers withdrew from the financial system. According to
Ellison, Whyley and Forster (2010), this act of withdrawal may be as a result of lack of trust,
costs, bad credit records, inappropriate product design and difficulties in managing spending. It is
also noteworthy that there is a huge difference between financial access and usage. While the
financially excluded can be encouraged to open an account, many of these accounts may become
dormant. Financial access therefore does not necessarily translate into usage for day-to-day
transactions (Platt et al. 2011).

In the literature there is no consensus on the determinants of financial inclusion. In a cross

country analysis, Sarma & Pais (2011) showed that income, income inequality, telephone and
internet usage and adult literacy are significant factors for financial inclusion in a country. They
contend that countries with low GDP per capita have comparatively poorer connectivity and
lower rates of literacy and seem to be more financially exclusive. In China, Fungáčová & Weill
(2015) showed that better education and higher income are correlated with higher usage of
formal accounts and formal credit. In Argentina, Tuesta, et al (2015) found that income and
education are all significant factors for financial inclusion. In India, Chithra & Selvam (2013)
found that income, population, literacy, deposit and credit penetration are significantly associated
with financial inclusion. As well, Kumar (2013) found that the socio‐economic and
environmental structure is significant in shaping the banking habits of the masses in India. In
Peru, Camara, Peña & Tuesta (2014) showed that income levels and education are significant
variables for the level of financial inclusion. In Africa, Allen et al. (2014) showed that population
density is highly more significant for financial inclusion than elsewhere. Besides, they found that
mobile banking expands financial access.

7
 O. Evans and B. Adeoye

Further recent work on financial inclusion has shown how Islamic bank impacts on financial
inclusion in Muslim countries, in particular Naceur, Barajas, & Massara (2015) and Demirguc-
Kunt, Klapper, and Randall (2013). Naceur, et al (2015) analyzed existing country-level data on
the relationship between Islamic banking and financial inclusion in Muslim countries.
Surprisingly they found that, though financial access has increased rapidly in these countries,
financial usage has not increased as quickly. Their evidence shows a positive but weak link
between financial inclusion and credit to households as well as to firms for financing investment.
Demirguc-Kunt, Klapper, and Randall (2013), using micro-level data, find that the scale of the
Islamic finance industry has no relation to differences in financial inclusion between non-
Muslims and Muslims.

Materials & Methods

Data
This study uses annual data (2005 to 2014) on number of depositors with commercial banks (per
1,000 adults), GDP per capita, deposit interest rates, inflation, money supply (% of GDP),
population, credit to the private sector (% of GDP), number of internet users, secure internet
servers, and adult literacy rate. Inevitably, data availability limits the data span to 2005-2014 and
the attention of the empirical analysis to 15 African economies, namely, Algeria, Botswana,
Cameroon, Angola, Ghana, Namibia, Niger, South Africa, Morocco, Kenya, Mali, Libya,
Malawi, Nigeria and Senegal.

Data for the analysis is collected from the World Development Indicators (WDI). The WDI is the
most appropriate source of macro-economic data, considering the cost of collecting primary data
for a large-scale panel study covering most of Africa. Also, WDI was chosen because it provides
an extensive range of information on the variables, better than what is obtainable in the Global
Financial index database on financial inclusion.

Consistent with the literature, the proxy for financial inclusion is the number of depositors with
commercial banks (per 1,000 adults) (Varman, 2005; Čihák, Demirgüç-Kunt, Feyen & Levine,
2012; Naceur et al, 2015). As well, the selected determinants of financial inclusion, consistent
with the literature, are broad money, deposit interest rate, domestic credit provided by financial

8
 O. Evans and B. Adeoye

sector as a % of GDP, GDP per capita, inflation, population, adult literacy, internet users per 100
people, secure internet servers and a dummy variable for Islamic banking presence and activity
(See Marshall, 2004; Sarma & Pais, 2011; Laha, Kuri & Kumar, 2011; Mohieldin, Iqbal, Rostom
& Fu, 2011; Demirguc-Kunt, Klapper & Randall, 2014; Mehrotra & Yetman, 2015; Naceur et al,
2015).

Model
The aim of this study is to determine the determinants of financial inclusion in Africa. In line
with existing studies (i.e. Marshall, 2004; Sarma & Pais, 2011; Laha, Kuri & Kumar, 2011;
Mohieldin, Iqbal, Rostom & Fu, 2011; Demirguc-Kunt, Klapper & Randall, 2014; Mehrotra &
Yetman, 2015; Naceur et al, 2015), the econometric model for the study is given as:

FINCit   0   1GDPCit   2 M 2GDPit   3CREDITit   4 INTERESTit   5 INFLATIONit   6 LITERACYit

  7USERS it   8 SERVER it   9 POPULATIONit   10 ISLAMICit   it (1)

Where FINC is financial inclusion (number of depositors with commercial banks per 1,000
adults); GDPC is GDP per capita; M2GDP is money supply (% of GDP) and CREDIT is the
credit to the private sector (% of GDP). INFLATION is headline inflation, USERS is the number
of internet users, SERVER is secure internet servers, LITERACY is adult literacy rate, and
POPULATION is the total population. INTEREST is the deposit interest rate. ISLAMIC is a
dummy variable which takes 1 if the country has Islamic banking presence and activity, and 0
otherwise.  are the residuals. The subscript i is the i-th country and the subscript t the t-th year.

Econometric Techniques
This study first carries out a panel unit root test on the variables in order to ascertain their
stationarity and preclude the possibility of spurious regression. Though the dynamic panel
approach is effective irrespective of whether the regressors are exogenous or endogenous, and
irrespective of whether the variables are I (0) or I (1) (Pesaran & Smith, 1995; Pesaran, Shin &
Smith, 1999), it is necessary to ensure that none of the variables is I(2). The Im Pesaran & Shin
(IPS, 2005) test for individual unit root process is the panel unit root test adopted for this study
and is given as:

pi
yit  i yi ,t 1  iL yi ,t L  zit   uit
L1
9
 O. Evans and B. Adeoye

(2)

The IPS test is appropriate for this study, considering the countries are heterogeneous. The IPS
test assumes the unit root can differ across the cross sections in the model. In other words, the
IPS test establishes a panel unit root test for the joint null hypothesis that individual series in the
model is non-stationary.

In order to adequately capture the dynamic processes between financial inclusion and its
determinants, this study uses the dynamic panel approach. According to Baltagi (2005, pp. 135),
“Many economic relationships are dynamic in nature and one of the advantages of panel data is
that they allow the researcher to better understand the dynamics of adjustment.” The benefit of
using the dynamic panel data model in this study is to introduce dynamic effects into the usual
panel data model (Baltagi, 1995); capture the dynamic effects of current or past shocks into the
model (Hsiao, 1986); control for both unobserved and missing variables or relationships; and
allow for identification of country-specific effects (Arellano-Bond, 1991; Pesaran, Smith, Im,
Matyas & Sevestre, 1996). The dynamic panel specifications in this study permits a high degree
of cross-country heterogeneity. This accounts for the fact that the determinants of financial
inclusion could vary across countries, contingent on country-specific structural factors such as
legal and institutional framework.

If yit is the dependent variable in country i, and xit is the vector of country-specific regressors
(Hsiao, 2003), then a modest dynamic panel data model can be set up as follows:

yit  yi ,t 1  xit  it [i=1,2,...,N; t=1,2,...,T] (3)

δ is a scalar, μi is the ith individual effect. The uit is a one-way error component model explained
by:

uit = µ i + νit [µ i ∼IID(0,σµ 2); νit ∼IID(0,σν2 ] (4)

µ i and νit are independent of each other and among themselves (Baltagi, 2005).

µ i is a vector of unobserved common factors.

10
 O. Evans and B. Adeoye

To allow for dynamics and cross-sectional dependence and slope heterogeneity in modelling the
relationship between financial inclusion and the determinants in a panel context, the dynamic
specification can be enhanced as follows:

p
y it   0   y y i ,t l    x x jit   it (5)
j 1

However, as a result of the inclusion of the lagged dependent variable yt-1in the model, “the
dynamic panel data regression is characterized by two sources of persistence over time:
autocorrelation due to the presence of a lagged dependent variable among the regressors and
individual effects characterizing the heterogeneity among the individuals” (Baltagi, 2005, pp.
135).While it has been established in the literature that this problem could hinder the robust
estimation of the model, a number of estimation techniques (i.e. Arellano & Bond (1991),
Arellano & Bover (1995), and Blundell & Bond (1998) using the generalized method of
moments (GMM estimator) has been developed to resolve the lagged dependent variable
problem in the panel setting (Deaton, 1997).

The facility to remove the across-time heterogeneity from equation (3.6) by taking first
differences is one of the greatest benefits of the GMM estimator in dynamic panel models
estimation:
p p
yit  yi ,t i    il ( yi ,t 1  yi ,t 2 )    il' ( xit  xi ,t l )  (  it   i ,t 1 ) (6)
l 1 l 0

|Using the Arellano-Bond estimator, higher lagged values of the dependent variable and the
exogenous regressors from all t periods can be used as instruments for the individual-specific
effects, (yt-1 - yt-2). On the contrary, Arellano & Bover (1995) showed that, to remove the
unobserved heterogeneity, the lagged dependent variable and explanatory variables (without first
differencing) and the lagged first differences can be used as instruments in the presence of time-
varying regressors uncorrelated with the country-specific effects. That is, in model 6, ∆yt-1 and xt-
1, xt-2, ..., x can be instruments for yt-1 and subsequently x can serve as instrument for y . The
instruments ensure that the GMM estimator gives consistent estimates.

11
 O. Evans and B. Adeoye

The consistency of dynamic panel approach demands sufficiently long lags, but longer lags than
necessary lead to estimates with very poor small sample properties (Elhorst, 2014). In this study,
the same lag order, 3, is used for all countries/variables, bearing in mind that a lag order of 3
should adequately account for the short-run dynamics. The likelihood of data mining is
precluded by the use of the same lag across all countries/variables. Note that the aim of this study
is to determine the determinants of financial inclusion in Africa rather than the country-specific
dynamics relevant to individual countries.

In order to ensure robustness of the results of this study, the validity tests of the instruments used
in the GMM estimation will be carried out (a scenario whereby the instruments are correlated
with the error process makes the validity of the instruments questionable). One of such tests is
Arellano & Bond’s (1991) specification test for lack of second-order serial correlation in the
first-difference residuals. The second specification test is the Sargan’s test of over-identifying
restrictions. To check the validity and the robustness of our results, therefore, the two tests are
employed.

Results & Discussion

Unit Root Test Result

The results of the IPS unit root test, as shown in Table 1, indicate that the variables are a mix of
I(0) and I(1) which is valid for the dynamic panel data approach. None of the variables is I(2).
Thus, we can safely begin the dynamic panel data estimation.

Table 1. IPS Unit Root Test

I(0) I(1) Decision

FINC 3.864 -2.059* I(1)

M2GDP 0.538 -1.920** I(1)

INTEREST -0.418 -2.349* I(1)

GDPC 2.838 -2.696* I(1)

INFLATION -1.982** -4.372* I(0)

12
 O. Evans and B. Adeoye

POPULATION 0.453 -2.389* I(1)

CREDIT 0.524 -3.605* I(1)

LITERACY 1.329 -2.648* I(1)

INTERNET -1.430 -1.953* I(1)

SERVERS -0.389 -3.491** I(1)

Source: Authors’ calculation using STATA 11

Notes: By Schwarz criterion, the lag length was 1. (*) and (**) indicate stationarity at
significance levels 1% and 5% respectively.
Having established that the variables are a mix of I(0) and I(1), it must be noted that the dynamic
approach is valid irrespective of whether the variables are I (0) or I (1), and irrespective of
whether the regressors are exogenous or endogenous (Pesaran & Smith, 1995; Pesaran & Shin,
1999; Pesaran, 1997).

Dynamic Panel Estimation

Depositors with commercial banks per 1,000 adults as a measure of financial inclusion (FINC) is
regressed on GDP per capita (GDPC), broad money (MONEY), deposit interest rate
(INTEREST), and domestic credit provided by financial sector as a % of GDP (CREDIT), and
internet users per 100 people (USERS), secure internet servers (SERVERS), inflation
(INFLATION), total population (POPULATION), adult literacy rate (LITERACY),and the a
dummy variable for Islamic banking presence and activity (ISLAMIC). This is necessary in
order to examine the contemporaneous effect of these variables on financial inclusion (FINC).
The Least Squares estimates obtained are thus reported for two cases2:
(a) Arellano-Bond dynamic panel-data and,
(b) Arrelano-Bover/Bundell-Bond system dynamic panel-data.
Table 4.2 shows the results of the dynamic panel estimation using both Arrelano-Bond and
Arrelano-Bover/Bundell-Bond methods. The coefficients on the lagged FINC are of special
interest in the setting of these two dynamic models. The lagged FINC estimates which are
statistically significant mean that lagged financial inclusion has significant impact on
contemporaneous financial inclusion and would thus indicate a “catch-up effect.” A coefficient
2 Individual country estimates are available on request, but take note they are likely to be independently undependable in view of
the fact that the time dimension of the panel is small.

13
 O. Evans and B. Adeoye

equal to zero would imply full catch-up, and a coefficient between zero and one would imply
partial catch-up, which is the case in our Arrelano-Bond Dynamic Panel and the Arrelano-
Bover/Blundell-Bond System dynamic panel models. The fact that the lagged financial inclusion
estimates are between zero and one implies that countries with stunted financial inclusion tend to
recover most of any financial inclusion deficit experienced in the past.
Table 2. Dynamic Panel Estimates
Arrelano-Bond Arrelano-Bover/Bundell-Bond
Coef. P>|z| Coef. P>|z|
Lagged FINC 0.530* 0.000 0.743* 0.000
GDPC 0.335** 0.040 0.317** 0.047
MONEY 98.484* 0.001 6.907* 0.000
CREDIT 0.674 0.263 0.659*** 0.088
INTEREST 5.117 0.178 5.106 0.175
INFLATION -0.691 0.392 -0.605 0.318
LITERACY 10.090** 0.046 8.152*** 0.073
POPULATION 17.814 0.492 15.378 0.411
USERS 3.960*** 0.057 3.957*** 0.050
SERVERS 0.046 0.149 0.049** 0.044
ISLAMIC 81.522* 0.000 79.003* 0.000
N = 105 N = 120
Wald χ2 = 456.91* Wald χ2 = 4308.880*
Sargan test = 8.536 Sargan test = 5.071
AB test = -0.025 AB test = -0.149
Source: Authors’ calculation using STATA 11.
Notes: The (*) signifies variable significant at 1%; (**) significance at 5%; (***) significance at
10%. test is Arellano and Bond test for AR(2). The Sargan test reports that under the null the
overidentified restrictions are valid.

GDPC is statistically significant and positive across both specifications. This means countries
with high per capita income have highly inclusive financial systems. This finding is in line with
Sarma & Pais (2011), Chithra & Selvam (2013), Camara et al. (2014), Tuesta, et al. (2015) and

14
 O. Evans and B. Adeoye

Fungáčová & Weill (2015) who also found that income is a significant variable for financial
inclusion in a country.

M2GDP is significant and positive across both specifications. Additionally, CREDIT is positive
but insignificant. The insignificant impact is expected, considering the fact that credit is
extremely low in Africa, due to a host of variables such as lack of collateral and credit
information. Nonetheless, this finding is in contrast with Chithra & Selvam (2013) who showed
that deposit and credit penetration have significant impacts on financial inclusion in India.

Population, though positive, is insignificant. This finding conflicts with Chithra & Selvam
(2013) who showed that population has significant impact on financial inclusion in India and
Allen et al. (2014) who also showed that population has significant impact on financial inclusion
in Africa. The impact of population on financial inclusion may have been overstated by these
studies. Inflation has a negative impact on the level of financial inclusion, though insignificant
across both specifications.

The deposit interest rate has positive but insignificant impacts on financial inclusion. The low
deposit interest rates in Africa are unlikely to significantly impact both existing and potential
depositors. Since the official interest rates is often the gauge of other interest rates in the
economy, broader access to financial services across Africa is likely to make the interest rates set
by African central banks a more potent device for regulating economies. In other words,
considering that the rewards for saving are influenced by interest rates, higher financial access
bring a bigger share of economic activity under the control of interest rates, making them a more
powerful tool for policymakers, but can as well worsen the risk of injurious financial crises.

Positive significant effects on financial inclusion are also seen, by way of literate rate. Literacy,
especially financial literacy, has gradually become more important as financial markets become
increasingly complex and the illiterate finds it difficult to make informed financial decisions.
This evidence is consistent with Sarma & Pais (2011) who, in a cross-country analysis, showed
that adult literacy is a significant factor in explaining the level of financial inclusion in a country
and Chithra & Selvam (2013) who found that literacy is an important in explaining the level of

15
 O. Evans and B. Adeoye

financial inclusion in India. Additionally, Camara et al (2014) and Tuesta et al (2015) showed
that better education is a significant variable for financial inclusion.

Internet users per 100 people (USERS) and secure internet servers (SERVERS) have significant
impacts on financial inclusion. This result is similar to Sarma & Pais (2011) and Allen et al.
(2014) who showed that internet access is an essential factor in a fast-moving and digital
economy. This is evidenced in the case of Kenya where M-Pesa has transformed a wide
spectrum of financial services. The significant impacts of internet access have very important
implications for financial inclusion. Without the intensive use of the internet in Africa, financial
inclusion will be very infinitesimal. Covering all the millions of villages in the African continent
with brick and mortar branches of financial institutions would be a very arduous task, in terms of
the investment and cost effectiveness. The internet has drastically reduced the cost of
transactions, via the mobile and the ATM. Further, internet has increased the potentials of credit
delivery in remote areas of the African continent. It has made it possible to provide home
banking services where the accounts are operated by illiterate customers using mobiles. The
internet, therefore, has become a major financial inclusion enabler.

The Islamic dummy variable is significant and positive across both specifications. In other
words, countries with Islamic banking presence and activity have higher financial inclusion. This
means that Sharia-compliant finance is an important factor for explaining the level of financial
inclusion. This result is consistent with Naceur et al (2015) who showed some evidence that
Islamic banking presence and activity is linked to higher financial inclusion in Muslim countries-
members of the Organization for Islamic Cooperation.

Conclusion

In this study, we have combined the Arrelano-Bond and Arrelano-Bover/Bundell-Bond dynamic

panel data approaches to assess the determinants of financial inclusion in 15 African countries.
This study finds that GDP per capita, broad money as a % of GDP, adult literacy rate, internet
access and Islamic banking presence and activity are significant factors explaining the level of
financial inclusion in Africa. Domestic credit provided by financial sector as a % of GDP,
deposit interest rates, inflation and population have insignificant impacts on financial inclusion

16
 O. Evans and B. Adeoye

in Africa. This study has highlighted the major financial inclusion-inducing factors, which may
help to improve future policy vis-à-vis financial inclusion.

However, while the findings of this study should be of help to African central banks’
policymakers and commercial bankers as they advance innovative approaches to enhance the
involvement of excluded poor people in formal finance, this study is far from an evaluation of
the financial inclusion drive in individual countries. The diversity of the countries in Africa
implies that the challenges encountered in one country may be quite different from the next.

There are a few critical areas for further research. Firstly, while the present study used the
number of depositors with commercial banks per 1,000 adults as a measure of financial
inclusion, it would be worthwhile to examine other alternative measures which could enhance
access to formal finance for excluded individuals, such as the nature and frequency of
transactions that take place in these accounts. Access is not synonymous with usage, and as such,
opening bank accounts without accompanying consistent usage may simply cause additional
costs for banks with no feasible advantage to poor African communities. Thus, future policy
measures to increase financial inclusion in Africa must give incentives for usage.

References

ABU‐BADER, S., & ABU‐QARN, A. S. (2008). Financial development and economic growth:
empirical evidence from six MENA countries. Review of Development Economics, 12(4),
803-817.
ADEOYE, B. W. (2015), Exchange rate Volatility and Inclusive Growth: The Nigerian
Experience, A paper presented at the 56th Annual Conference of the Nigerian Economic
Society (NES) with the theme “Attaining Inclusive Growth in Nigeria: Prospects and
Challenges” held at Sheraton Abuja Hotel between October 11 and 14.

ADEOYE, B. W. & M. O. SAIBU (2014), “Monetary Policy Shocks and Exchange Rate
Volatility in Nigeria”, Asian Economic and financial Review, 4(4): 544 – 562.

17
 O. Evans and B. Adeoye

ADEOYE, B. W. & SANGOSANYA, A.O. (2015), “Monetary Policy and Inclusive Growth:
The Case of Nigeria”, A paper presented at the 56th Annual Conference of the Nigerian
Economic Society (NES) with the theme “Attaining Inclusive Growth in Nigeria:
Prospects and Challenges” held at Sheraton Abuja Hotel between October 11 and 14.

AGA, G.A. & MARTINEZ PERIA, M.S., (2014). International remittances and financial
inclusion in Sub-Saharan Africa. World Bank Policy Research Working Paper, (6991).
AGHION, P., & PATRICK B. (1997). A Theory of Trickle-Down Growth and Development.
Review of Economics Studies 64: 151–72.
ALLEN, F., CARLETTI, E., CULL, R., SENBET, L., & VALENZUELA, P. (2014). The
African financial development and financial inclusion gaps. Journal of African
economies, eju015.
APORTELA, F. (1999). Effects of Financial Access on Savings by Low-Income People. MIT
Department of Economics Dissertation Chapter 1.
ARELLANO, M., & BOND, S. (1991). Some tests of specification for panel data: Monte Carlo
evidence and an application to employment equations. The review of economic studies,
58(2), 277-297.
ARELLANO, M., & BOVER, O. (1995). Another look at the instrumental variable estimation of
error-components models. Journal of econometrics, 68(1), 29-51.
ARYA, P. K. (2015). Financial Inclusion and their role in inclusive growth. International
Journal of Management, IT and Engineering, 5(2), 79.
ASHRAF, N., D. KARLAN, AND W. YIN. 2010. Female Empowerment: Further Evidence
from a Commitment Savings Product in the Philippines. World Development 28 (3): 333–
44.
AYYAGARI, M. AND BECK, T., (2015). Financial Inclusion in Asia: An Overview. Asian
Development Bank Economics Working Paper Series, (449).
BALTAGI, B. (2008). Econometric analysis of panel data (Vol. 1). John Wiley & Sons.
BALTAGI, B. H. (1995). The econometrics of panel data. WestSussex: Wiley.
BANERJEE, ABHIJIT, AND ANDREW NEWMAN, 1993. Occupational Choice and the
Process of Development. Journal of Political Economy 101: 274–98.

18
 O. Evans and B. Adeoye

BARAJAS, M.A., CHAMI, M.R. AND YOUSEFI, M.R., (2013). The finance and growth nexus
re-examined: Do all countries benefit equally? (No. 13-130). International Monetary
Fund.

BECK, T., A. DEMIRGÜÇ-KUNT AND P. HONOHAN. (2009). Access to financial services:

Measurement, impact, and policies. The World Bank Research Observer 24(1): 119-145.
BECK, T., A. DEMIRGÜÇ-KUNT, AND R. LEVINE. 2010. “Financial Institutions and
Markets across Countries and over Time: The Updated Financial Development and
Structure Database.” World Bank Economic Review 24 (1): 77–92.
BEN NACEUR, S., BARAJAS, A., & MASSARA, A. (2015). Can Islamic Banking Increase
Financial Inclusion? (No. 15/31). International Monetary Fund.
BHATTACHARYA, R. & H. WOLDE, 2012, Constraints on Growth in the MENA Region,
IMF Working Paper 10/30 Washington: International Monetary Fund.
CAMARA, N., PEÑA, X., & TUESTA, D. (2014). Factors that matter for financial inclusion:
Evidence from Peru (No. 1409).

CHAKRABORTI, S., & SANYAL, S. (2015). Heuristic Algorithm using Internet of Things and
Mobility for solving demographic issues in Financial Inclusion projects. arXiv preprint
arXiv:1503.00369.
CHITHRA, N., & SELVAM, M. (2013). Determinants of financial inclusion: An empirical study
on the inter-state variations in India. Available at SSRN 2296096.

ČIHÁK, M., DEMIRGÜÇ-KUNT, A., FEYEN, E., & LEVINE, R. (2012). Benchmarking
financial systems around the world. World Bank Policy Research Working Paper, (6175).
DE KOKER, LOUIS AND JENTZSCH, NICOLA (2011), Financial inclusion and financial
integrity : aligned incentives?, in Shadow 2011 : The shadow economy, tax evasion and
money laundering: Proceedings of the 2011 Shadow conference, University of Münster,
Münster, Germany, pp. 1-28.
DEATON, A. (1997). The analysis of household surveys: a microeconometric approach to
development policy. World Bank Publications.

19
 O. Evans and B. Adeoye

DEMETRIADES, P., & ANDRIANOVA, S. (2004). Finance and growth: what we know and
what we need to know. In Financial Development and Economic Growth (pp. 38-65).
Palgrave Macmillan UK.

DEMIRGÜÇ-KUNT, A & L KLAPPER (2012): Measuring financial inclusion: the Global

Findex database, World Bank, Policy Research Working Paper, no 6025.

DEMIRGUC-KUNT, A., KLAPPER, L., & RANDALL, D. (2014). Islamic finance and financial
inclusion: measuring use of and demand for formal financial services among Muslim
adults. Review of Middle East Economics and Finance, 10(2), 177-218.
DEMIRGUC-KUNT, A., KLAPPER, L. AND RANDALL, D., (2014). Islamic finance and
financial inclusion: measuring use of and demand for formal financial services among
Muslim adults. Review of Middle East Economics and Finance, 10(2), pp.177-218.
DITTUS, P. & M. KLEIN (2011). On Harnessing the Potential of Financial Inclusion, BIS
Monetary and Economic Department. Working Papers. No 347. BIS, Basel.
DUPAS, P., & J. ROBINSON. (2009) Savings Constraints and Microenterprise Development:
Evidence from a Field Experiment in Kenya. National Bureau of Economic Research
Working Paper 14693.
ELHORST, J. P. (2014). Spatial econometrics: from cross-sectional data to spatial panels.
Heidelberg, New York, Dordrecht, London: springer.
ELLISON, A., C. WHYLEY & R. FORSTER (2010). Realising Banking Inclusion: The
Achievements and Challenges. Financial Inclusion Task Force. HM Treasury, London.
ESCHENBACH, F. (2004) Finance and Growth: A Survey of the Theoretical and Empirical
Literature. SSRN Working Paper Series.
EUROPEAN COMMISSION (2008). Financial Services Provision and Prevention of Financial
Exclusion. Directorate-General for Employment, Social Affairs and Equal Opportunities,
Brussels.

FEYEN, E., K KIBUUKA AND D. SOURROUILLE, 2013, “FinStats 2014: A ready-to-use tool
to benchmark financial sectors across countries and time” World Bank mimeo
(Washington).

FINSCOPE (2004). South Africa 2004 – Survey Highlights. FinScope, Midrand.

20
 O. Evans and B. Adeoye

FUNGÁČOVÁ, Z., & WEILL, L. (2015). Understanding financial inclusion in China. China
Economic Review, 34, 196-206.
GALOR, O., & ZEIRA, J. (1993). Income distribution and macroeconomics. The review of
economic studies, 60(1), 35-52.
GOLDSMITH, R. (1969). Financial structure and economic development. New Haven: Yale
University Pres.
GOODHART, C. A. E. (2004). Some new directions for financial stability? (Vol. 27). Bank for
International Settlements.
IM, K. S., PESARAN, M. H., & SHIN, Y. (2003). Testing for unit roots in heterogeneous panels.
Journal of econometrics, 115(1), 53-74.
JENTZSCH, N. (2011). Implications of Mandatory Registration of Mobile Phone Users in
Africa, Mimeo, DIW Berlin.
JOHNSTON, R. B., & SUNDARARAJAN, V. (1999). Sequencing Financial Sector Reforms.
International Monetary Fund.
KUMAR, N. (2013) Financial inclusion and its determinants: evidence from India, Journal of
Financial Economic Policy, 5, 4 – 19

LAHA, D., KURI, D., & KUMAR, P. (2011). Determinants of financial inclusion: A study of
some selected Districts of West Bengal, India. Indian Journal of Finance, 5(8), 29-36.
LOVE, I. & MARTINEZ PERIA, M.S., (2012). How bank competition affects firms' access to
finance. World Bank Policy Research Working Paper, (6163).
MARSHALL, J. N. (2004). Financial institutions in disadvantaged areas: a comparative analysis
of policies encouraging financial inclusion in Britain and the United States. Environment
and Planning A, 36(2), 241-261.
MCKINNON, R. I. (1973). Money and capital in economic development. Brookings Institution
Press.
MEHROTRA, A. N., & YETMAN, J. (2015). Financial Inclusion-Issues for Central Banks. BIS
Quarterly Review March.
MIRDALA, R. (2011). Financial deepening and economic growth in the European transition
economies. Journal of Applied Economic Sciences, 6(2), 177-194.

21
 O. Evans and B. Adeoye

MOHIELDIN, M., IQBAL, Z., ROSTOM, A. M., & FU, X. (2011). The role of Islamic finance
in enhancing financial inclusion in Organization of Islamic Cooperation (OIC) countries.
World Bank Policy Research Working Paper Series, Vol.

NACEUR, M. S. B., BARAJAS, M. A., & MASSARA, M. A. (2015). Can Islamic Banking
Increase Financial Inclusion? (No. 15-31). International Monetary Fund.

NGUENA, C. L., & ABIMBOLA, T. M. (2014). Financial Deepening Dynamics and Implication
for Financial Policy Coordination in WAEMU. Regional Integration and Policy
Challenges in Africa, 201.
OHWOFASA, B. O., & AIYEDOGBON, J. O. C. Financial Deepening and Economic Growth in
Nigeria, 1986-2011: An Empirical Investigation.
PESARAN, H., SMITH, R., IM, K., MATYAS, L., & SEVESTRE, P. (1996). The econometrics
of panel data: A Handbook of the theory with applications.
PESARAN, M. H., & SMITH, R. (1995). Estimating long-run relationships from dynamic
heterogeneous panels. Journal of econometrics, 68(1), 79-113.
PESARAN, M. H., SHIN, Y., & SMITH, R. P. (1999). Pooled mean group estimation of
dynamic heterogeneous panels. Journal of the American Statistical Association, 94(446),
621-634.
PLATT, A-B., A. SINGH, S. BANSAL, A. GIRI, AND A.J. TIWARI (2011). No Thrills –
Dormancy in NFA Accounts. Microsave India, Lucknow.
SALAHUDDIN, M. (2015). Ethical Banking System: An Alternate For Financial Inclusion.
Abhinav-National Monthly Refereed Journal Of Research In Commerce & Management
(Online ISSN 2277-1166), 4(3), 12-15.
SARKAR, S. (2015). Financial Inclusion and Rural Development–The Role of India Post. Asian
Journal of Research in Social Sciences and Humanities, 5(4), 185-194.
SARMA, M., & PAIS, J. (2011). Financial inclusion and development. Journal of international
development, 23(5), 613-628.

SHERRADEN, M. S., & GRINSTEIN‐WEISS, M. (2015). Creating Financial Capability in the

Next Generation: An Introduction to the Special Issue. Journal of Consumer Affairs,
49(1), 1-12.

22
 O. Evans and B. Adeoye

TRIKI, T., & FAYE, I. (2013). Financial Inclusion in Africa. Tunisia: African Development
Bank.
TUESTA, D., SORENSEN, G., HARING, A., & CAMARA, N. (2015). Financial inclusion and
its determinants: the case of Argentina (No. 1503).
VARMAN, M. (2005). Impact of self-help groups on formal banking habits. Economic and
Political Weekly, 1705-1713.
WOLDE-RUFAEL, Y. (2009). Re-examining the financial development and economic growth
nexus in Kenya. Economic Modelling, 26(6), 1140-1146.

WORLD BANK (2014), Financial Inclusion, Global Financial Development Report

(Washington).

ZHUANG, J., GUNATILAKE, H. M., NIIMI, Y., KHAN, M. E., JIANG, Y., HASAN, R, &
HUANG, B. (2009). Financial sector development, economic growth, and poverty
reduction: A literature review. Asian Development Bank Economics Working Paper
Series, (173).

23